San Francisco’s ChatTJB presents itself as AI, but every reply is written by its founder

A new chatbot project in San Francisco called ChatTJB is drawing attention for presenting an apparent twist on AI: there is no actual artificial intelligence behind it. According to BlockBeats, the service was created by artist and former Google employee Tucker Bryant, who personally reads, considers, and manually answers every user prompt. On its website, ChatTJB describes itself as 「Artisanal Intelligence」 and a 「Single-Operator Large Language Experience」, or LLE. Bryant said the project is meant to satirize users’ growing dependence on AI chat tools. In his view, as products such as ChatGPT become more common, people are increasingly willing to accept answers that sound confident and polished while doing less independent thinking. Bryant said ChatTJB intentionally sacrifices efficiency so users can experience waiting, uncertainty, and the process of thought again. He added that the project is not anti-AI, but an attempt to get people to think again about how they consume machine-generated responses. The service can also respond to image-generation requests with hand-drawn pictures. Its website states clearly that ChatTJB is a satirical public art project rather than a real AI product.